Nonexperimental methods, such as regression modeling or statistical matching, produce unbiased estimates if the underlying assumptions hold, but these assumptions are usually not testable. Most studies testing nonexperimental designs find that they fail to produce unbiased estimates, but these...

Using data from Mathematica's experimental evaluation of charter schools, this methodological study examines the validity of four different comparison group approaches to test whether these designs can replicate findings from a well-implemented random assignment study.

Using data from Mathematica's experimental evaluation of charter schools, this methodological study examines the validity of four different comparison group approaches to test whether these designs can replicate findings from a well-implemented random assignment study.

Nonexperimental methods, such as regression modeling or statistical matching, produce unbiased estimates if the underlying assumptions hold, but these assumptions are usually not testable. Most studies testing nonexperimental designs find that they fail to produce unbiased estimates, but these...

Examines children's diets and finds nearly all children have adequate usual intakes for most B vitamins, but many children of all ages are at risk of inadequate intakes of folate, magnesium, zinc, and vitamins A and E. In addition, a large proportion of children have calcium intakes well below...

Summarizes the largest federal study ever conducted of dropout prevention programs, finding three promising models: alternative middle and high schools for those at risk of dropping out, programs aimed at helping older students earn a GED, and whole-school restructuring efforts aimed at dropout...